H A R R Y J. G E N S L E R

A K A N T I A N A R G U M E N T A G A I N S T A B O R T I O N *

(Received 14 December, 1984)

I f you asked ten years ago for my view on the morality of abortion, I would

have said ' ! don ' t have a view - the issue confuses me'. But now I think

that abortion is wrong and that certain Kantian consistency requirements

more or less force us into thinking this. Part III will present my reasoning.

But first, in Parts I and II, I will show why various traditional and recent

arguments on abortion do not work.

I. A T R A D I T I O N A L A N T I - A B O R T I O N A R G U M E N T

One common traditional argument goes this way:

The killing of innocent human life is wrong.

The fetus is innocent human life.

.'. The killing o f the fetus is wrong.

This seemingly simple argument raises some difficult questions:

Is it 'always wrong' or 'normally wrong'? And if the latter, how

do we decide the difficult cases?

Is the fetus ' innocent ' if it is attacking the life or health or social

well-being of the woman?

Is there a clear and morally-weighty distinction between 'killing'

and 'letting die' - or between 'direct killing' and 'indirect killing'?

I will not discuss these important questions; a short article on abortion must

leave many questions unanswered. But I will discuss this one: 'What does the

* This aztiele appeared originally in Philosophical Studies 48, pp. 57-72, and is reprinted here because of some pages having appeared out o f their proper order.

Philosophical Studies 49 (1986) 83-98. �9 1986 by D. Reidel Publishing Company

84 H A R R Y J . G E N S L E R

term "human life" in the abort ion argument m e a n ? ' People sometimes

presume that the meaning o f the term is clear and that the major problem is

the factual one of whether the fetus is 'human life' (in some clear sense). But

I think that the term in this context is fuzzy and could be used in different

senses.

Suppose we found a Martian who could discuss philosophy; would he be

'human'? We need to make distinctions: the Martian would be 'human' in

the sense o f 'animal capable o f reasoning' ( ' rat ional animal ' ) but not in the

sense of 'member o f the species h o m o sapiens' - so the Martian is 'human'

in one sense but not in another . Which of these senses should be used in the

abort ion argument? The fetus is not yet an 'animal capable of reasoning'. Is it a 'member of the species h o m o sapiens'? That depends on whether the un-

born are to be counted as 'members ' of a species - ordinary language can use

the term either way. In the biology lab we all (regardless of our views on

abort ion) distinguish between 'human' fetuses and 'mouse ' fetuses - so in

this sense (the 'genetic sense') the fetus is human. But in counting the number

of mice or humans in the city of Chicago we all (regardless of our views on

abort ion) count only the born - so in this sense ( ' the populat ion-study

sense') the fetus is not a human. So is the fetus a 'human'? In two senses o f

this term that we have distinguished the answer would be 'NO' while in a

third sense the answer would be 'YES'; whether the fetus is 'human' depends

on what is meant by 'human' .

Human life has been claimed to begin at various points:

(1) at conception.

(2) when individuality is assured (and the zygote cannot split or fuse

with another).

(3) when the fetus exhibits brain waves.

(4) when the fetus could live apart.

(5) at birth.

(6) when the being becomes self-conscious and rational.

Here we do not have a factual disagreement over when there emerges, in the

same clear sense of the term, a 'human' ; rather we have six ways to use the

term. Answer (1) is correct for the 'genetic sense', ( 5 ) f o r the 'populat ion-

study sense', and (6) for the ' rational animal sense'; answers (2) to (4) reflect

other (possibly idiosyncratic) senses. And there are likely other senses of

'human' besides these six. Which of these senses are we to use in the first

premise ( 'The killing of innocent human life is wrong')? We get different

A K A N T I A N A R G U M E N T A G A I N S T A B O R T I O N 85

principles depending on which sense of the term 'human ' we use.

Can we decide which sense to use by appealing to scientific data? No, we

cannot. Scientific data can help us judge whether a specific individual is

'human' in some specified sense (e.g. sense (3) or sense (4)) but it cannot tell

us which sense of 'human ' to use in our principle.

Can we decide by ' intui t ion ' - by following the principle that seems most

correct? Note that moral intuit ions depend greatly on upbringing and social

milieu. Most Catholics were brought up to have intuitions in line with sense

(1) (the 'genetic sense'). Many ancient Romans and Greeks were trained to

have sense (6) intuit ions (allowing abort ion and infanticide). And many

Americans today are being brought up to have sense (5) intuit ions (allowing

abort ion but not infanticide). Is there any way to resolve this clash - other

than simply praising our own intuit ions and insulting contrary ones? Can we

carrry on the argument further? I think we can and that the Kantian appeal

to consistency provides a way to resolve the issue rationally.

I1. SOME R E C E N T P R O - A B O R T I O N A R G U M E N T S

Before getting to the Kantian approach, let us consider three arguments in

defense of abort ion. A common utilitarian argument goes this way:

Anything having a balance of good results (considering everyone)

is morally permissible.

Abor t ion often has a balance of good results (considering every-

o n e ) .

.'. Abor t ion often is morally permissible.

Here 'good results' is most commonly interpreted in terms of pleasure and

pain ( 'hedonist ic act uti l i tarianism') or the satisfaction of desires ( 'prefer-

ence act util i tarianism').

The second premise (on the good results of abor t ion) is controversial.

People defending the premise say that abort ion often avoids difficulties

such as the financial burden o f a child on poor parents or on society, the

disruption o f schooling or a career, and the disgrace of an unwed mother;

that where these problems or probable birth defects exist, the child-to-be

would have less chance for happiness; and that abor t ion provides a 'second

chance' to prevent a birth when contraceptives fail or people want to rethink

an earlier choice. But opponents say that we can have equally good results

without abort ion, by using better social structures (more social support to-

86 H A R R Y J . G E N S L E R

ward unwed mothers and poor families, better adoption practices, wiser

use of contraceptives, etc.) and scientific advances (better contraceptives,

artificial wombs, etc.); and they say that abortion can harm the woman

psychologically and promote callous attitudes toward human life.

I think the weaker link is the first premise - the argument's utilitarian

basis. This premise would often justify killing, not just fetuses, but also in-

fants and the sick or handicapped or elderly; many utilitarian reasons for not

wanting a child around the house would also apply to not wanting grand-

mother around. And the premise would justify these killings, not just when

they have great utilitarian benefits, but even when the utilitarian benefits are

slight. Utilitarianism says that the killing of an innocent human being is

justified whenever it brings even a slight increase in the sum-total of pleasure

(or desire-satisfaction). This is truly bizarre. Imagine a town where lynchings give the people great pleasure (or satisfy

their desires) and the ultilitarian sheriff lynches an innocent person each week

because the pleasure (or desire) of the masses slightly outweighs the misery

(or frustration of desire) of the person to be lynched - and so the action has

a slight gain in 'good results'. If the utilitarian principle is correct then the

sheriff's lynchings are morally justfied! But could anyone really believe that

these lynchings would be morally justified? I could pile up further examples of strange and unbelievable implications

of utilitarianism. Utilitarians try to weasel out of these examples -- but I

think not with ultimate success. So my verdict on utilitarianism is that it would justify so many bizarre actions (including so many killings) that we

would not accept this principle if we were consistent and realized its logical

consequences. My second pro-abortion argument is from Michael Tooley. 1 Tooley recog-

nizes that humans have a right to life presumably a greater right than

utilitarians would recognize; but only humans in sense (6) ('rational ani- mals' - or, as he puts it, 'persons') have such a right. The human fetus, while it might develop into a being with a right to life, presently has no more right to

life than a mouse fetus. A fetus lacks a right to life because 'rights' connect

with 'desires' conceptually - so that you can have rights only if you have

desires. Tooley's argument is roughly this:

A being has a right to X only if it desires X.

No fetus desires its continued existence [because then the fetus

A KANTIAN ARGUMENT AGAINST ABORTION 87

would have to have a concept of itself as a continuing subject of

experiences - a concept it cannot as yet have].

.'. No fetus has a right to its continued existence.

Tooley claims that the first premise is not correct as it stands; we must add

three qualifications to make the premise harmonize with our intuitions re-

garding rights:

A being has a right to X only if either it desires X or else it would

desire X were it not (a) emotionally unbalanced or (b) tem-

porarilly unconscious or (c) conditioned otherwise.

He thinks the revised first premise will serve equally well (assuming obvious

changes in the second premise); so he concludes that fetuses (and infants) do

not have a right to life.

But we need further exceptions to make the first premise correspond to our intuitions. If we think that the dead have rights (e.g. to have their wills

followed), then we need to add 'or (d) the being did desire X when it was

alive'. I f we think that a child who lacks the concept 'hepatitis' (and thus

cannot desire not to be given this disease) does not thereby lose his right not to be given hepatitis, then we need to add 'or (e) the being would desire X if

it had the necessary concepts'. If we think (as I do) that trees and canyons

have the right not to be destroyed without good reason, then we would have

to add some exception for this. And if we think that the fetus (or infant) has

a right to life, then we need to add something like 'or (f) if the being were to

grow up to be an adult member of the rational species to which it belongs

then it would desire to have had X ' (presumably if the fetus were to grow up

to be an adult member of homo sapiens then it would desire to have had

continued life - and this, with (O, allows the fetus to have a right to life). 2

The trouble with Tooley's argument is that disagreements over the main

issue of the right to life of the fetus translate into disagreements over how to

qualify the first premise to make it mesh with ~our' intuitions; so the argu-

ment cannot decide the main issue.

The third argument in defense of abortion comes from Judith Jarvis

Thomson and presumes that the fetus is a 'person' (in some undefined sense): 3

One who has voluntarily assumed no special obligation toward

another person has no obligation to do anything requiring great

personal cost to preserve the life of the other.

88 H A R R Y J. G E N S L E R

Often a pregnant woman has voluntarily assumed no special obligation toward the unborn child (a person), and to preserve

its life by continuing to bear the unborn child would require

great personal cost.

.'. Often a pregnant woman has no obligation to continue to bear the unborn child.

The first premise here seems acceptable. Normally you have no obligation to

risk your life to save a drowning stranger; if you risk your life then you do more than duty requires. But it is different if you are a lifeguard who has as-

sumed a special obligation - then you have to try to save the person, even at

the risk of your own life. Thomson thinks that a woman getting pregnant intending to have a child is voluntarily accepting a special obligation toward

the child. However if the pregnancy is accidental (the result of a contracep-

tive failure or rape) then the woman has assumed no such special obligation

and, if continuing to bear the child requires great personal cost, the woman

has no obligation to continue to bear it, the woman would do no wrong if she has an abortion - but if she continues to bear the child in spite o f personal

cost then she is doing something heroic, something beyond what duty requires.

Thomson gives an analogy. Suppose you wake up and find yourself in bed

with an unconscious violinist attached to your circulatory system (his friends

attached him to you because this was needed to save his life); if you discon-

nect him before nine months, he will die - otherwise he will live. Even

though it might be praiseworthy to make the sacrifice and leave him plugged

in for nine months, still you have no obligation to do so; it would be morally

right for you to disconnect him, even though he will die. So also if you are

pregnant under the conditions mentioned above, then, even though it might

be praiseworthy to make the sacrifice and bear the child for nine months,

still you have no obligation to do so; it would be morally right for you to

have the child removed, even though it will die.

The first premise of Thomson's argument is slightly misstated. A motorist

has a special obligation toward a person he has injured in an accident, even

though he has not voluntarily assumed this obligation in any clear way (the

accident happened against his will and despite all reasonable precautions - just like an accidental pregnancy). Similarly a child has a special obligation to- wards its parents - even though he has not voluntarily assumed this obliga- tion. Not all special obligations toward others are 'voluntarily assumed' - so

A K A N T I A N A R G U M E N T A G A I N S T A B O R T I O N 89

these two words should be crossed out in the premises.

My main objection to the argument can be put as a dilemma. Utilitarianism is either true or false. If it is true, then the first premise is false (because then

the person has an obligation to do whatever has the best consequences -

despite personal cost); and so the pro-abortion utilitarian Peter Singer rejects

this premise, since it conflicts with utilitarianism. But if utilitarianism is false,

then presumably Sir David Ross was right in claiming it to be morally signifi-

cant that others:

... stand to me in relation of promisee to promiser, of creditor to debtor, of wife to husband, of child to parent [my emphasis], of friend to friend, of fellow countryman to fellow countryman, and the like; and each of these relations is the foundation of a prima facie duty, which is more or less incumbent on me according to the circumstances of the case. 4

If utilitarianism is false, then likely a person has greater obligations toward his

or her offspring than toward a violinist stranger - and so the second premise,

which claims that the pregnant woman has no special responsibility toward

her own child, begins to look doubtful (recall that we crossed out the words

'voluntarily assumed').

I I I . A K A N T I A N A R G U M E N T

My Kantian approach to abortion stresses consistency. In discussing utilitari-

anism I appealed to simple logical consistency (not accepting a principle

without accepting its recognized logical consequences). Here I will use two

further consistency requirements (based on the universalizability and

prescriptivity principles) and a third consistency requirement derived from

these two (a version of the golden rule). The following argument displays these three requirements and how the third follows form the first two:

If you are consistent and think that it would be all right for

someone to do A to X , then you will think that it would be all

right for someone to do A to y o u in similar circumstances.

I f you are consistent and think that it would be all right for some- one to do A to you in similar circumstances, then you will con-

sent to the idea of someone doing A to you in similar circum-

stances.

.'. If you are consistent and think that it would be all right to do A

90 H A R R Y J . G E N S L E R

to X, then you will consent to the idea of someone doing A to you in similar circumstances. (GR)

The first premise can be justified by the 'universalizability principle', which

demands that we make similar ethical judgments about the same sort of

situation (regardless of the individuals involved); so if I think it would be all

right to rob Jones but I don ' t think it would be all right for someone to rob

me in an imagined exactly similar situation, then I violate universalizability

and am inconsistent. The second premise can be justified by the 'prescrip-

tivity principle', which demands that we keep our ethical beliefs in har-

mony with the rest o f our lives (our actions, intentions, desires, and so

forth); so if I think an act would be all right but I don ' t consent to it being

done, then I violate prescriptivity and am inconsistent. These and further

derived requirements can be formulated and justified in a rigorous way; but

I won' t do that here. The conclusion GR is a form of the golden rule; if I

think it would be all right to rob Jones but yet I don' t consent to (or

approve of) the idea o f someone robbing me in similar circumstances, then I violate GR and am inconsistent, s

The following argument combines an instance of GR with an empirical premise about your desires:

If you are consistent and think that stealing is normally permis- sible, then you will consent to the idea of people stealing from you in normal circumstances. (From GR)

You do not consent to the idea of people stealing from you in

normal circumstances.

:. If you are consistent then you will not think that stealing is

normally permissible.

Most of us do not consent to the idea of people stealing from us in normal

circumstances; so we would not be consistent if we held 'Stealing is normally

permissible' (since then we would violate consistency principle GR). This argument shows that, given that a person has a certain desire (one that most

people can be presumed to have), he would not be consistent if he held a given ethical view. The conclusion here concerns the consistency of holding

the ethical judgment and not the judgment's truth. A person could escape

this conclusion if he did not care if people robbed him; then the second premise would be false. Throughout the rest of this article I will

A K A N T I A N A R G U M E N T A G A I N S T A B O R T I O N 91

generally assume that the reader desires not to be robbed or blinded or

killed; if you would love people to rob or blind or kill you (or you don' t care whether they do this to you) then most of my further conclusions

will not apply to you. It might seem easy to argue similarly on abortion. How would you like it

ff someone had aborted you? Should we say that you don' t like the idea and

so you can't consistently hold that abortion is permissible? Or should we say that as an ignorant fetus you would not have known enough to have been

against the abortion - so that this argument won' t work? Let us slow down and try to understand GR more clearly before applying it

to abortion. Properly understood, GR has to do with my present reaction toward a hypothetical case - not with how I wouM react if I were in the

hypothetical case. A few examples may clarify things. Consider this chart:

Issue

Do I think it permis-

sible to rob X while X

is asleep?

Right Question

Do I now consent to the

idea o f my being robbed

while asleep?

Wrong Question

If I were robbed while

I was asleep would I

then (while asleep) con-

sent to this action?

(In the 'Right Question' and 'Wrong Question' I presume implicit 'in rele-

vantly or exactly similar circumstances' qualifiers). The point of this chart

is that, by GR, to be consistent in answering YES to the ISSUE I must also

answer YES to the RIGHT QUESTION - but I need not answer YES to the

WRONG QUESTION. Presumably I would answer NO to the RIGHT QUES-

TION; when I consider the hypothetical case o f my-being-robbed-while- asleep I find that I now (while awake) do not consent to or approve of this

action. But the WRONG QUESTION has to do with what I, ff I were robbed while asleep, would consent to or approve of while thus asleep (and thus

ignorant o f the robbery); GR, correctly understood, has nothing to do with

the WRONG QUESTION. Let me give another example:

Issue

Do I think it permis- sible to violate X ' s will

after his death?

Right Question

Do I now consent to the idea of my will being

violated after my death?

Wrong Question

If my will is violated

after my death, would I then (while dead) con-

sent to this action?

92 H A R R Y J . G E N S L E R

Again GR has to do with my present reaction toward a hypothet ical case

in which I may imagine myself as asleep or dead or even a fetus - but not

with how I would react while asleep or dead or a fetus in the hypothet ical

situation.

But is it legitimate to apply the golden rule to our t reatment of a fetus?

Consider a case not involving abort ion:

Issue

Do I think it permis-

sible to bl ind X while

X is a fetus?

Right Question

Do I now consent to

the idea of my having

been blinded while a

fetus?

Wrong Question

If I were bl inded while

a fetus, would I then

(while a fetus) consent

to this action?

Suppose that you had a sadistic mother who, while pregnant with you, con-

templated injecting herself with a blindness-drug which would have no effect

on her but which would cause the fetus (you) to be born bl ind and remain bl ind all its (your) life. Your mother could have done this to you. Do you

think this would have been all r i g h t - and do you consent to the idea of her

having done this? The answer is a clear NO - and an equally clear NO regard-

less o f the t ime o f pregnancy that we imagine the injection taking place. We

could then argue as we did concerning stealing:

If you are consistent and think that blinding a fetus is normally permissible, then you will consent to the idea of your having been blinded while a fetus in normal circumstances. (F rom GR)

You do not consent to the idea of your having been blinded while

a fetus in normal circumstances.

.'. I f you are consistent then you will not think that blinding a fetus

is normally permissible.

Again, with most people the second premise will be true - most people can

be presumed not to consent to (or approve of) the idea of this act having

been done to them.

Is it legitimate to apply the golden rule to our t reatment of a fetus? Surely

it is - the above reasoning makes good sense. If a pregnant woman is about

to do something harmful to the fetus (like taking drugs or excessive alcohol

A K A N T I A N A R G U M E N T A G A I N S T A B O R T I O N 93

or cigarettes), it seems appropriate for her to ask, 'How do I now react to the

idea o f my mother having done this same thing while she was pregnant with

me?' Applying the golden rule to a fetus raises no special problems.

But someone might object as follows:

Seemingly your view forces us to accept tha t the fetus has fights (e.g. no t to be blinded by the drug), even though you avoid saying it is human . But your quest ion about "my having been blinded while a fetus' presupposes that the fetus and my present self are identical - the same human being. So aren't you presupposing (despite your earlier discussion on the many senses of 'human') that the fetus is 'human'?

While my way of phrasing the question may presuppose this, I put my ques-

tion this way only for the sake o f convenience; I could re-phrase my ques-

tion so that it doesn't presuppose this:

Do I now consent to the idea of: - my having been blinded while a fetus?

- the fetus that developed into my present self having been

blinded?

- Helen E. Gensler having taken the blindness-drug while preg-

nant in 1945?

The second and third way to phrase the question do not presuppose that the

fetus and my present self are identical or the same human being; if you wish,

you may rephrase my comments thusly (I will keep to the first way of

speaking for the sake of brevity). I am against the idea of the drug having

been given, not because I think that the fetus was in some metaphysical sense

the same human being as I, but rather because if this drug had been given then I would be blind all my life.

The application o f GR to abortion is similar - we need only switch from a blindness-drug (which blinds the fetus) to a death-drug (which kills the fetus).

Your mother could have killed you through such a death-drug (or other

means o'f abortion). Do you think this would have been all right - and do

you consent to (or approve of) the idea o f her having done this? Again the

answer is a clear NO - and an equally clear NO regardless of the time of

pregnancy that we imagine the killing taking place. We can argue as we did concerning blinding:

If you are consistent and think that abortion is normally per-

missible, then you will consent to the idea o f your having been

aborted in normal circumstances. (From GR)

94 H A R R Y J, G E N S L E R

You do not consent to the idea of your having been aborted in

normal circumstances.

- I f you are consistent then you will not think that abortion is

normally permissible.

Again with most people the second premise will be true - most people can

be presumed not to consent to (or approve of) the idea of this act having

been done to them. So insofar as most people take a consistent position they

will not think that abortion is normally permissible.

IV. SIX O B J E C T I O N S

(1) Surely a utilitarian would see your two drug cases as very different - the blindness-drug inflicts needless future suffering while the death-drug simply eliminates a life. Why wouldn ' t a utilitarian, moved by the greatest total happiness principle, approve of the death-drug having been given to him if this would have led to a greater total happiness? Wouldn ' t such a person be a consis tent upholder o f the view tha t abort ion is normally permissible?

My answer is that utilitarianism leads to so many strange moral implications

that, even if the utilitarian could be consistent on this one case, still he

would likely be inconsistent in his overall position. I previously claimed

that utilitarianism would justify so many bizarre actions (including so many killings) that we would not accept this principle if we were consistent and

realized its logical consequences. But ff there are few (if any) consistent

utilitarians then there would be few (if any) consistent utilitarian upholders

of the view that abortion is normally permissible.

(2) Let us consider a non-utilitarian who approves o f abort ion bu t not infanticide or the blindness-drug. Why couldn ' t such a person consent to the idea of himself having been aborted under imagined or actual normal circumstances - and hence be consistent?

Such a person could be consistent, but only with bizarre desires about how he himself is to be treated. Let us suppose that someone combined these

three judgments (as many are being brought up to do in our society today):

(a) It is wrong to blind an adult or child or infant or fetus. (b) It is wrong to kill an adult or child or infant. (c) It is permissible to kill a fetus.

A KAN T I A N A R G U M E N T A G A I N S T A B O R T I O N 95

To be cons i s t en t the pe r son wou ld have to answer these ques t ions as fol lows:

Do you consent to the idea of my blinding you now? - NO!

Do you consent to the idea of my having blinded you yesterday? - NO!

... when you were five years old? - NO!

... when you were one day old? - NO!

... before you were born? - NO!

Do you consent to the idea of my killing you now? - NO!

Do you consent to the idea of my having killed you yesterday? - NO!

... when you were five years old? - NO!

... when you were one day old? - NO!

... before you were born? - YES[I!

It is s t range t h a t the p e r s o n disapproves equally o f be ing blinded at t he

var ious t imes - and disapproves equally o f be ing killed at t he first f ou r

t imes - and yet approves o f be ing killed at the last t ime. He opposes the

b l ind ings because , regardless o f t he i r t iming, the e f fec t wou ld be the same -

he wou ld be b l ind. He opposes t he kil l ings at the first f ou r t imes because ,

again, the ef fec t w o u l d be the same - he w o u l d n o t be alive; b u t ki l l ing at the

f i f th t ime has the same ef fec t - w h y shou ld he no t oppose th i s ki l l ing also?

The 'YES ' here seems r a t h e r s t range. O f course one w h o t h i n k s his life n o t

w o r t h living could give a ' Y E S ' to the idea o f his having b e e n ki l led whi le a

f e tus - b u t t h e n we wou ld e x p e c t 'YES ' answers to the idea o f his be ing

ki l led at the o t h e r t imes as well ( w h i c h w o u l d m a k e h i m incons i s t en t i f he

he ld t h a t it is w r o n g to kill an adu l t or chi ld or in fan t ) . So whi le a non -

u t i l i t a r ian w h o c o m b i n e s the th ree j u d g m e n t s above couM in pr inc ip le have

such desires and be cons i s ten t , still th is is un l ike ly to h a p p e n very o f t e n - to

be cons i s t en t the p e r s o n wou ld have to have very b izar re desires. 6

(3) Are you saying that the desires that most people have are good while unusual (or 'bizarre') desires are bad? How would you establish this?

I am no t saying t h a t c o m m o n desires are good while unusua l desires are

bad - o f t e n the reverse is t rue ; and s o m e t i m e s w h e n we no t i ce a conf l i c t

b e t w e e n ou r m o r a l bel iefs and ou r desires we come to change ou r desires and

no t o u r m o r a l beliefs. R a t h e r I am appea l ing to desires t h a t m o s t peop le have

because I am t ry ing to deve lop a cons i s t ency a r g u m e n t to show t h a t m o s t

peop le w h o a d o p t the p r o - a b o r t i o n view are incons i s ten t . In e f fec t I am

chal lenging those w h o a d o p t such a view b y saying, ' L o o k at w h a t you wou ld

have to desire in o rde r to be cons i s t en t in y o u r p o s i t i o n - go and t h i n k a b o u t

it and see w h e t h e r you real ly are c o n s i s t e n t ! ' I c la im t h a t m o s t o f the t i m e

the p ro - abo r t i on i s t will f ind t h a t he is i ndeed incons i s t en t - he is s u p p o r t i n g

96 H A R R Y J. G E N S L E R

certain moral principles about the t reatment o f others that he would not wish

to have been followed in their actions toward him.

(4) You question the consistency of one who holds that abortion is permissible but infanticide is wrong. But let us see whether you are consistent. If it would have been wrong for your parents to have aborted you, wouldn't it have been equally wrong for your parents not to have conceived you? The result would have been the same - there would be no YOU!

My answer here is complicated. My first reaction is to disapprove of the idea

of m y parents not having conceived me - to think it would have been wrong

for them to have abstained or used contraceptives; but the universalizing

requirement forces me to change m y reactions (whereas it doesn' t do this in

the abor t ion case). If I hold 'I t is wrong to have an abort ion in this (my)

case', then I have to make the same judgment in all similar cases; but I can

easily hold (consistently) that it is in general wrong to have an abortion. But

if I hold ' I t is wrong to prevent conception (by e.g. abstinence or contracep-

tives) in this (my) case', then I again have to make the same judgment in all

similar cases; but I cannot hold (consistently) that it is in general wrong to

prevent conception - since this would commit me to desiring a policy which

would bring about a greatly overpopulated world of starving people at a very

low level of human life. So, in order to be consistent, I change my first

reaction and come to judge that it would have been morally permissible for

my parents not to have conceived (me) on August 5, 1944 - but instead

perhaps to have conceived (someone else) on September 5, 1944 - and I

come, though with hesitation, to consent to the possibility of their having

done this. To sum up: the universalizing requirement points to an important

difference between aborting and not conceiving - I can 'will as a universal

law' a general prohibi t ion against aborting, but not one against non-con- ceiving.

(5) Suppose that reason does force up into thinking that abort ion is normally wrong. What does 'normal ' here mean? And aren ' t the 'abnormal ' or 'unusual ' cases the more impor tan t and difficult ones to deal with? So isn' t your conclusion unimpor tan t?

My claim that abort ion is normally wrong means that it is wrong in at least the

great majori ty of cases but perhaps not in every conceivable case (e.g. in the

imagined case where Dr. Evil will destroy the world if we do not do an abor-

t ion). The question o f what unusual conditions ( if any) would just ify abor-

t ion is indeed important and difficult. But I think that, in light of the very

A KANTIAN A R G U M E N T AGAINST ABORTION 97

great n u m b e r o f ' conven ience abor t ions ' going on today , the issue o f the

general mora l s ta tus o f abor t ion is at the p resen t t ime far more impor t an t .

(6) Suppose that i f I am consistent I cannot hold that abortion is normally permissible. What if I do not care about being consistent? Can you prove to me that I ought to care? Or can you prove to me that abortion is wrong without appealing to consistency?

You ask too much . Suppose I give you an a rgument proving tha t abor t ion is

wrong (or tha t you ought to care abou t being consis tent ) . I f you do no t

already care abou t cons i s tency , w h y should you no t accept the premises o f

m y a rgument and ye t reject the conclus ion? This would be incons is ten t - bu t

you d o n ' t care abou t this! So you p resumably w o u l d n ' t care abou t any

a rgument I might give - in e f fec t you are saying tha t you have a closed mind .

I f you d o n ' t care abou t cons i s t ency t h e n I am wast ing m y t ime w h e n I t ry to

reason wi th you.

NOTES

i Tooley's original argument was in 'Abortion and infanticide', Philosophy and Public Affairs 2 (1972), pages 37-65. He added refinements to his view in Philosophy and Public Affairs 2 (1973), pages 419-32; in a postscript to a reprint of his article in The Rights and Wrongs of Abortion, edited by Marshall Cohen, Thomas Nagel, and Thomas Scanlon (Princeton, 1974), pages 80-84; and in 'In defense of abortion and infanticide', in The Problem o f Abortion (second edition), edited by Joel Feinberg (Belmont CA, 1984), pages 120-134. (The weak link in the latest version of the argument seems to be this premise: 'An individual existing at one time cannot have desires at other times unless there is at least one time at which it possesses the concept of a continuing self or mental substance'; this entails the incredible 'Your pet kitten cannot yesterday have had a desire to eat unless at some time it possesses the concept of a continuing self or mental sub- stance.') Peter Singer's defense of abortion and infanticide rests partially on Tooley's earlier argument but mainly on his preference utilitarianism; see chapters 4 and 6 of his Practical Ethics (Cambridge, 1979). 2 Clause (f) was phrased to skirt the issue of Tooley's "superkittens" who become rational if given a certain drug; my intuitions on the superkitten (and Frankenstein) cases are not very clear. Clause (f) may require further refinement. 3 'A defense of abortion', in Philosophy and Public Affairs 1 (1971), pages 47 -66. 4 The Right and the Good (Oxford, 1930), page 19. 5 In arguing the abortion issue, 1 use some ideas from the theory of R. M. Hare, as developed in his Freedom and Reason (Oxford, 1963). Hare once wrote an article on 'Abortion and the golden rule' (Philosophy and Public Affairs 4 (1975), pages 201-22); but his approach differs from mine. Hare rests his case on 'We should do to others what we are glad was done to us' and on the fact that we are glad that we were conceived, not aborted, and not killed as infants; hence we too ought to conceive, not abort, and not kill infants (but contraception, abortion, and infanticide turn out to have only a weak prima facie wrongness which is easy to override by other considerations). Hate's formulation of the golden rule here is defective; if I am glad my parents gave me hundreds

98 H A R R Y J . G E N S L E R

o f gifts each Christmas, then pe rhaps to be consis tent I mus t hold tha t it would be good to do this same thing in similar circumstances - bu t I net~d not hold tha t one should do this ( that it is a duty). Also m y conclusions differ f rom Hate 's - I view abort ion and infanticide (but no t failing-to-conceive) as seriously wrong; I th ink m y conclusions are what Hare's theory should lead to.

On the Tooley/Singer view the cut-off point for killing is no t bir th bu t rather when the child comes to desire its cont inued existence as a cont inuing subject o f experiences. (It is unclear at what age this happens.) My response to this view would be m u c h like the above, except tha t the killing side o f the chart would now have one more YES. (I would like to thank m y colleague Paul Moser for some helpful suggestions on the first draft o f this paper.)

Department o f Philosophy, Loyola University, 6525 North Sheridan, Chicago, IL 60626, U.S.A.